Aquiles-ai/Athenea-4B-VL-Thinking
LoRA-style vs huihui-ai/Huihui-Qwen3-VL-4B-Thinking-abliterated. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 36 of 36 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/4, Norms 0/249, MLP 108/220, Attention 144/240
2 derivatives checked
- mradermacher/Athenea-4B-VL-Thinking-i1-GGUF: Quantization (fewer layers)
- mradermacher/Athenea-4B-VL-Thinking-GGUF: Same weights, converted (fewer layers)